Editor’s Note: Welcome to our weekly Reader Forum section. In an attempt to broaden our interaction with our readers we have created this forum for those with something meaningful to say to the wireless industry. We want to keep this as open as possible, but we maintain some editorial control to keep it free of commercials or attacks. Please send along submissions for this section to our editors at: [email protected].
It is apparent that we are currently experiencing an explosion in mobile data traffic and that wireless operators globally are rapidly increasing wireless network capacity both in the wireless portion of the network and in the backhaul. Now, as operators plan for 3G and LTE migration, they are looking ahead and planning backhaul networks that will cope with the even large volume of data traffic that this will bring. In many cases this involves migrating to fiber-based backhaul and “fibering up” cell sites.
It is often said that the most expensive part of deploying any new fiber network is digging the holes in the ground to run new ducts so naturally this creates an excellent business opportunity for fiber owners to service wireless operators with mobile backhaul services. Operators looking to provide these wholesale backhaul services will find that having the right fiber in the right place can be a key asset.
Building a transport network through synchronous Ethernet
To capitalise on this opportunity, wholesale wireline operators and cable MSOs must consider how best to build a transport network over this fiber asset. The market place offers a broad range of solutions for building optical and packet-optical networks, each with varying levels of economics and scalability. Naturally each vendor will claim to provide the lowest cost solution initially and throughout the life of the network. These cost models are best assessed by simple modelling exercises with a shortlist of vendors. But, in addition to ensuring the best whole-life economics, there are some key technical features to be considered that can have a significant impact on the service quality offered; and therefore, the competitiveness in the mobile backhaul market.
Beyond the obvious features, such as ultra-low latency and jitter to support LTE backhaul, operators should also consider network synchronization. Current synchronization techniques vary considerably across the globe with a preference for GPS-based synchronisation in North America and for E1-based network synchronisation in Europe and Asia. However, as operators move to higher capacity backhaul and consider the need for smaller cell sites for LTE (which may be located where GPS may not be suitable, such as in shopping malls and subway stations), wireless operators are reassessing their synchronization schemes and future plans.
Networks globally are starting to move to Ethernet-based backhaul using synchronous Ethernet (SyncE) or timing-over-packet schemes, such as 1588v2. Adoption of these technologies varies between wireless operators and wholesale backhaul providers need to keep their options open by providing both. If they choose the right implementation of SyncE enabled backhaul, where the synchronization is created by clocking the Ethernet network itself in a similar manner to SONET or SDH networks, then they also have the opportunity to offer significantly better quality synchronization than was achievable over SONET/SDH based networks. These SONET/SDH networks were previously considered to be the gold-standard for synchronized networks. In the market today, there are wholesale providers who supply their wireless operator customers with synchronization that is more than 10-times better than they previously had over E1/SDH based synchronisation. This improvement has tangible benefits for network performance and ultimately customer satisfaction.
Flexibility, performance and scalability
Another benefit of SyncE backhaul is the use of SyncE synchronization domains. A backhaul network may be required to support multiple wireless technologies with differing clocks or even multiple wireless operators. Supporting transparent wholesale services with SyncE over a transport network isn’t always easy and not all optical systems can support them, let alone multiple SyncE signals over the same wavelength. However, some do and this capability can have a significant impact on the economics of supporting multiple wireless technologies or multiple customers over the same backhaul infrastructure.
Let’s now consider 1588v2 as the alternative to SyncE for synchronization over an Ethernet service. Here the timing information is encoded into sync data packets and sent over the network without the Ethernet network being synchronized itself. This is potentially easier for a wholesale operator to support as they don’t necessarily need to support the functionality in all nodes, just in the source and destination nodes that perform the ingress and egress functions where the synchronization data is inserted or extracted from Ethernet frames. Often the wireless operators will perform this within their own equipment and they will simply need the Ethernet backhaul network to provide a reliable and stable transport of these frames. This makes things considerably easier for the wholesale operator but certain factors determine how the 1588v2 mechanism can scale. These primarily relate to latency, jitter and packet loss in the Ethernet network. The better the underlying performance of the wholesale Ethernet backhaul service the better the performance and scalability of the 1588v2 synchronization.
Wholesale operators should not have to dictate synchronization mechanisms or the size of synchronization domains to their wireless operator customers due to limitations within their network. If at all possible, they should provide the wireless operator with a solution that keeps their options open and allows them the flexibility to use any synchronization mechanism in their network. If wholesale operators do not do this, then they may find that a competitor does, suddenly leaving them completely exposed if the wireless operator wants the alternative mechanism or scale that they aren’t able to support.
Real-world implementation
An excellent example of a wholesale and business services operator who has recently moved into wireless backhaul is Virgin Media Business in the United Kingdom. Virgin Media Business already provides wholesale services for core transport to a number of mobile operators in the United Kingdom, but had previously not offered backhaul services from the cell site back to the core of the wireless operators’ networks.
The new service has already attracted it’s first customer, MBNL (Mobile Broadband Network) which is the joint venture backhaul company for 3 (Hutchinson Whampoa) and Everything Everywhere which is the recently merged U.K. operations of Orange (France Telecom) and T-Mobile (Deutsche Telekom). This 8-year deal has a contract value of approximately $150 million and provides an uncontended gigabit Ethernet service to a large number of cell sites across the United Kingdom spread across 14 regional aggregation networks.
A key factor in Virgin Media’s win was the ability to provide exceptionally high quality network synchronization using the SyncE capabilities of the packet optical technology they chose to deploy. With very low latency and virtually zero jitter, the network also provides excellent support for 1588v2 should the mobile operators choose to also run it over the network.
So, in addition to checking out the usual factors of economics, scalability and low-latency, wholesale operators who wish to capitalise on this new business opportunity should investigate the quality of their SyncE support, both in terms of layer two performance and multiple domains at layer one. Also they should dig deep in terms of layer two Ethernet performance and its impact on 1588v2 scalability. This will enable them to take the right approach to Ethernet transport, which can have a major impact on how they can differentiate their services in this very competitive market.
